Lovable valued at $6.6B as AI coding startup funding keeps climbing

CNBC reported the round on 16 December. It is Lovable's third this year. The sources, who asked to remain anonymous while discussing private information, said U.S. venture firm Accel took part. One of them also named Khosla Ventures as an investor. Neither the company nor the two funds had responded to CNBC's request for comment when the article was published.

That is a lot of money for a company founded in 2023.

Accel is not a new name here. The firm backed Lovable's previous round and has taken positions in several of the sector's largest financings, among them billion-dollar rounds for Cursor maker Anysphere and for Thinking Machines, the AI company founded by former OpenAI executive Mira Murati. Khosla Ventures had not previously been reported as a participant in this round. Forbes reported in November that the deal would value Lovable at around $6 billion, below the figure the two anonymous sources gave CNBC.

The growth numbers are what make the valuation legible. Lovable reported $200 million in annual recurring revenue in November, less than a year after it first hit $1 million in ARR. The company said at the time that 100,000 projects were being built on its platform every day. Its July fundraise brought in $200 million and drew investors including Creandum, Klarna founder Sebastian Siemiatkowski, ElevenLabs founder Mati Staniszewski and Synthesia founder Victor Riparbelli.

What Lovable actually sells

Lovable is based in Stockholm and is opening offices in Boston and San Francisco.

Its platform uses AI models from providers including OpenAI and Anthropic to let users build apps and websites through text prompts, without needing to understand code. That is the core of what the industry has taken to calling vibe coding, and it is where a large share of this year's venture money has gone. The comparison set is expensive. Anysphere raised $2.3 billion at a $29.3 billion valuation in November, according to CNBC. Replit reached a $3 billion valuation in September after picking up $250 million, and Vercel closed a $300 million round at a $9.3 billion valuation. Those are US companies. Lovable is the European leader in the category, which is part of why its valuation matters beyond its own balance sheet.

Software that writes software is not the only place capital is concentrating.

Edtech tells the opposite story. Global investment in education technology peaked at $16.7 billion in 2021, when lockdowns kept millions of children out of classrooms, then fell to less than $3 billion by 2025, according to Tracxn, a Bengaluru-based platform that tracks startup funding. Rest of World reported the figures in April 2026. The number of new edtech companies launched fell from almost 10,500 in 2020 to 645 in 2025.

The money did not disappear. It moved.

HolonIQ, a research firm that advises governments and investors on education, wrote in a February post that investors had concentrated capital in AI-enabled products, workforce-aligned platforms and K-12 operations tools that address cost or operational pressures. Loot Drop, a database of more than 1,700 startup closures, concluded that corporate workforce development and specialised skill acquisition for high-income careers look more promising than K-12 general education. Byju's, once the world's most valuable education startup at $22 billion, collapsed under a financial crisis and aggressive sales tactics. Nigerian startup Edukoya shut down in 2025.

The risk that arrives after the money

One lawsuit filed this year shows what can go wrong when AI-generated output is published as fact.

MeetingTV, a video conferencing and webinar startup, sued Palo Alto Networks after its newly acquired Koi Security threat-intelligence business published a blog linking the startup to a Chinese corporate espionage operation. The Register reported the complaint on 2 July 2026. The filing alleges Koi used a large language model to generate the threat report, that the AI system hallucinated findings about MeetingTV, and that the security shop published those as facts on 30 December.

Koi's blog, which was later edited silently to remove references to MeetingTV's Zoomcorder product, had described the meeting recording service as a "public-facing front" for a Chinese criminal operation. The lawsuit says the report caused security companies and service providers worldwide to block MeetingTV's domains and services, labelling them malware and command-and-control infrastructure.

"If people on the internet are blocked from reaching your company, then that's a death sentence," MeetingTV founder and CEO Michael Robertson told The Register.

Robertson said Koi never contacted his company before publishing. He said he learned about the report only when providers began blocking him, and that Verizon and Palo Alto Networks were still blocking the startup. A Palo Alto Networks spokesperson told The Register the company was aware of the lawsuit regarding a report published by Koi before the acquisition, and said it believed Koi's research reflected its commitment to identifying and exposing threats. The spokesperson declined to answer specific questions about MeetingTV's allegations.

The case is not a funding story. It is the reason some investors are asking harder questions about what AI products actually produce, and who checks the output before it ships.

Meanwhile, at the other end of the market, the experimentation continues cheaply. A GitHub project called Mini-AGI, posted to Hacker News in September 2026, describes a continual learning byte-level language model that assembles its own architecture, trains from scratch on a single 8 GB VRAM GPU and stores its weights as ordinary files on disk, paging them onto the card as needed. Its author is explicit that it is a toy-level model and not a frontier system. The weights have not been published. The run is still reading its first pass over the corpus.

That is a small experiment. Lovable's round is the scale of the bet being placed on the other side of the same idea.